Predicting Costs Associated with Open Heart Surgery Based on Clinical, Administrative and Cost Data

Abstract

Cardiovascular disease is a leading cause of mortality and morbidity in the United States. Coronary artery bypass graft (CABG) surgery, a leading revascularization procedure to treat coronary artery disease, is a high cost procedure that results in large economic impact. Ability to predict CABG surgery cost could enable clinicians and administrators to better manage hospital resources. To plan for CABG surgery cost based on individual patient characteristics, this study develops predictive models using clinical, administrative and cost data. We applied semiparametric regression to develop (i) a cost model that consists of pre-operative variables, and (ii) a cost model that consists of pre-, peri- and post-operative variables. Adding perioperative and postoperative variables increased model accuracy by 25%. Statistically significant variables can inform clinicians and administrators to focus on areas for quality and process improvement initiatives. Potential limitation for model adoption is that costing methodology and accounting methods might vary across hospitals.